Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y V O L . 6 7 , N O . 2 3 , 2 0 1 6

ª 2 0 1 6 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O UN DA T I O N I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 6 . 0 3 . 5 2 4

Ticagrelor for Prevention of IschemicEvents After Myocardial Infarction inPatients With Peripheral Artery Disease

Marc P. Bonaca, MD, MPH,a Deepak L. Bhatt, MD, MPH,a Robert F. Storey, MD,b Ph. Gabriel Steg, MD,c

Marc Cohen, MD,d Julia Kuder, MS,a Erica Goodrich, MS,a José C. Nicolau, MD, PHD,e Alexander Parkhomenko, MD,f

José López-Sendón, MD,g Mikael Dellborg, MD,h Anthony Dalby, MD,i Jind�rich �Spinar, MD,j Philip Aylward, MD,k

Ramón Corbalán, MD,l Maria Teresa B. Abola, MD,m Eva C. Jensen, MD, PHD,n Peter Held, MD, PHD,n

Eugene Braunwald, MD,a Marc S. Sabatine, MD, MPHa

ABSTRACT

Fro

Inf

dio

Sch

Pa

tig

GokD

vis

Me

BACKGROUND Peripheral artery disease (PAD) is associated with heightened ischemic and bleeding risk in patients

with prior myocardial infarction (MI).

OBJECTIVES This study evaluated the efficacy and safety of ticagrelor on major cardiovascular (CV) events and major

adverse limb events in patients with PAD and a prior MI.

METHODS PEGASUS-TIMI 54 (Prevention of Cardiovascular Events in Patients With Prior Heart Attack Using Ticagrelor

Compared to Placebo on a Background of Aspirin—Thrombolysis In Myocardial Infarction 54) randomized 21,162 patients

with prior MI (1 to 3 years) to ticagrelor 90 mg twice daily, ticagrelor 60 mg twice daily, or placebo, all on a background of

low-dose aspirin. History of PAD was obtained at baseline. Occurrences of major adverse cardiovascular events (MACE)

(defined as CV death, MI, or stroke) and major adverse limb events (MALE) (defined as acute limb ischemia or peripheral

revascularization for ischemia) were recorded in follow-up.

RESULTS A total of 1,143 patients (5%) had known PAD. In the placebo arm, those with PAD (n ¼ 404) had higher rates

of MACE at 3 years than those without (n ¼ 6,663; 19.3% vs. 8.4%; p < 0.001), which persisted after adjusting for

baseline differences (adjusted hazard ratio: 1.60; 95% confidence interval: 1.20 to 2.13; p ¼ 0.0013), and higher rates of

acute limb ischemia (1.0% vs. 0.1%) and peripheral revascularization procedures (9.15% vs. 0.46%). Whereas the

relative risk reduction in MACE with ticagrelor was consistent, regardless of PAD, patients with PAD had a greater ab-

solute risk reduction of 4.1% (number needed to treat: 25) due to their higher absolute risk. The absolute excess of TIMI

major bleeding was 0.12% (number needed to harm: 834). The 60-mg dose had particularly favorable outcomes for CV

and all-cause mortality. Ticagrelor (pooled doses) reduced the risk of MALE (hazard ratio: 0.65; 95% confidence interval:

0.44 to 0.95; p ¼ 0.026).

CONCLUSIONS Among stable patients with prior MI, those with concomitant PAD have heightened ischemic risk. In

these patients, ticagrelor reduced MACE, with a large absolute risk reduction, and MALE. (Prevention of Cardiovascular

Events in Patients With Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin [PEGASUS-

TIMI 54]; NCT01225562) (J Am Coll Cardiol 2016;67:2719–28) © 2016 by the American College of Cardiology Foundation.

m the aTIMI Study Group, Brigham and Women’s Hospital Heart & Vascular Center, Boston, Massachusetts; bDepartment of

ection, Immunity and Cardiovascular Disease, University of Sheffield, Sheffield, United Kingdom; cFrench Alliance for Car-

vascular Trials, Université Paris-Diderot, Paris, France; dCardiovascular Division, Newark Beth Israel Medical Center, Icahn

ool of Medicine at Mount Sinai, New York, New York; eHeart Institute (InCor)–University of São Paulo Medical School, São

ulo, Brazil; fUkranian Strazhesko Institute of Cardiology, Kiev, Ukraine; gHospital Universitario La Paz, Instituto de Inves-

ación La Paz, Madrid, Spain; hDepartment of Molecular and Clinical Medicine, Sahlgrenska Academy, University of Gothenburg,

thenburg, Sweden; iLife Fourways Hospital, Randburg, South Africa; jUniversity Hospital, Jihlavska, Brno, Czech Republic;

ivision of Medicine, Cardiac & Critical Care Services, Flinders Medical Centre, South Australia, Australia; lCardiovascular Di-

ion, Pontificia Universidad Católica de Chile, Santiago, Chile; mPhilippine Heart Center, University of the Philippines College of

dicine, Manila, Philippines; and nAstraZeneca R&D, Mölndal, Sweden. The TIMI Study Group has received significant research

grant supp

Dr. Bhatt h

Bioscience

is chair of

Clinical Re

received h

Publication

Clinical Re

ology), Jou

trial steeri

Patient Car

served as v

Publication

Laboratorie

(Editor, Ca

Scientific,

FlowCo, PL

received p

Scientific.

Amarin, A

Janssen, M

personal fe

Janssen, M

grants from

support fro

and Sanofi

Novartis, a

received p

Boehringer

received a

received gr

of the stud

and as a sp

and has re

Institute. D

received in

and Wom

GlaxoSmith

Alnylam, A

that they h

Manuscrip

ABBR EV I A T I ON S

AND ACRONYMS

ALI = acute limb ischemia

CABG = coronary artery

bypass graft

CI = confidence interval

HR = hazard ratio

MACE = major adverse

cardiovascular event(s)

MALE = major adverse limb

event(s)

MI = myocardial infarction

PAD = peripheral artery

disease

TIA = transient ischemic attack

TIMI = Thrombolysis In

Myocardial Infarction

Bonaca et al. J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6

Ticagrelor in Patients With PAD and Prior MI J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8

2720

P atients with peripheral artery dis-ease (PAD) are at heightened risk ofmajor adverse cardiovascular events

(MACE), including myocardial infarction(MI) and stroke. The presence of concomitantsymptomatic PAD in patients with coronarydisease (i.e., polyvascular disease), has beenassociated with a further heightened risk ofischemic events and mortality beyond thatassociated with symptomatic disease ineither vascular territory (coronary or periph-eral artery) alone (1–4).

SEE PAGE 2729

Intensive antithrombotic strategies aimedat reducing risk of MACE in patients withPAD have shown mixed results. Whereasantiplatelet monotherapy has been shown to

ort from AstraZeneca. Dr. Bonaca has received consulting fees from

as served on the advisory board of Cardax, Elsevier Practice Up

s; has served on the Board of Directors of Boston VA Research Inst

the American Heart Association Quality Oversight Committee; has

search Institute, Harvard Clinical Research Institute, Mayo Clinic

onoraria from the American College of Cardiology (Senior Associa

s (Editor-in-Chief, Harvard Heart Letter), Duke Clinical Research

search Institute (clinical trial steering committee), HMP Commun

rnal of the American College of Cardiology (Guest Editor; Associate E

ng committee), Slack Publications (Chief Medical Editor, Cardiol

e (Secretary/Treasurer), and WebMD (CME steering committees); h

ice-chair of the NCDR-ACTION Registry Steering Committee; ha

s Committee; has received research funding from Amarin, Astra

s, Ischemix, Medtronic, Pfizer, Roche, Sanofi, and The Medicin

rdiovascular Intervention: A Companion to Braunwald’s Heart Di

and St. Jude Medical; is a trustee of the American College of Ca

x Pharma, and Takeda. Dr. Storey has received grants, personal

ersonal/consultancy fees from Aspen, PlaqueTec, Correvio, The M

Dr. Steg has received research grants from Merck, Sanofi, and Se

straZeneca, Bayer, Boehringer-Ingelheim, Bristol-Myers Squibb, C

erck, Novartis, Pfizer, Regeneron, Sanofi, Servier, and The Medi

es from AstraZeneca during the conduct of the study; outside of t

aquet, malpractice attorneys, Merck, Bristol-Myers Squibb/Pfizer

Janssen and Edwards Lifesciences. Dr. Nicolau has received

m Amgen, AstraZeneca, Bayer, Bristol-Myers Squibb, Boehringer

. Dr. López-Sendón has received research grants from AstraZen

nd Servier; and has received honoraria from AstraZeneca, Lilly

ersonal fees from AstraZeneca during the conduct of the study;

Ingelheim, DiscoveryHealth, Lilly, Novartis, Pfizer, Sanofi, and Se

research grant, has received honoraria for speaking, and served

ants andmodest personal/speaker fees from and served on the loc

y; as an investigator in other ticagrelor trials has received modest

eaker for Pfizer; has received modest personal fees as speaker fo

ceived modest investigator fees from Bayer, Boehringer Ingelhei

r. Jensen is an employee of AstraZeneca. Dr. Held is an employee

stitutional grant support from AstraZeneca. Dr. Sabatine has

en’s Hospital from Abbott Laboratories, Amgen, AstraZeneca,

Kline, Intarcia, Merck, Novartis, Poxel, Roche Diagnostics, Sano

mgen, AstraZeneca, Cubist, CVS Caremark, Intarcia, and Merck (

ave no relationships relevant to the contents of this paper to dis

t received February 23, 2016; revised manuscript received March

reduce ischemic risk in patientswith symptomatic PAD(4–6), the combination of aspirin and a P2Y12 inhibitorin patients with PAD undergoing lower extremitybypass surgery did not reduce a broad composite ofcardiovascular and limb endpoints (7). However, largerstudies evaluating the combination of aspirin andclopidogrel have shown trends of benefit for reducingMACE in patients with symptomatic vascular disease,including those with PAD (8,9). More recently, theaddition of a PAR-1 antagonist, vorapaxar, was shownto reduce the risk of MACE in patients with prior MI orPAD but increase GUSTO (Global Use of Strategies toOpen Occluded Arteries) moderate or severe bleeding(10). In addition to reducing MACE, vorapaxar wasshown to reduce limb morbidity, including acute limbischemia (ALI) and peripheral revascularization pro-cedures, when added to aspirin and/or clopidogrel(11,12). These findings highlight the potential of

AstraZeneca, Merck, Bayer, and Roche Diagnostics.

date Cardiology, Medscape Cardiology, and Regado

itute, and the Society of Cardiovascular Patient Care;

served on the data monitoring committees of Duke

, and the Population Health Research Institute; has

te Editor, Clinical Trials and News, ACC.org), Belvoir

Institute (clinical trial steering committees), Harvard

ications (Editor-in-Chief, Journal of Invasive Cardi-

ditor), Population Health Research Institute (clinical

ogy Today’s Intervention), Society of Cardiovascular

as served as deputy editor of Clinical Cardiology; has

s served as the chair of the VA CART Research and

Zeneca, Bristol-Myers Squibb, Eisai, Ethicon, Forest

es Company; has received royalties from Elsevier

sease); is a site coinvestigator for Biotronik, Boston

rdiology; and has performed unfunded research for

fees, and travel support from AstraZeneca; and has

edicines Company, Medscape, and Thermo Fisher

rvier; and has served as a consultant or speaker for

SL Behring, Daiichi-Sankyo-Lilly, GlaxoSmithKline,

cines Company. Dr. Cohen has received grants and

his study he has received personal fees from Merck,

, Boehringer Ingelheim, and Lilly; and has received

speaker/consulting honoraria and/or research grant

Ingelheim, GlaxoSmithKline, Merck, Novartis, Pfizer,

eca, Lilly/Daiichi-Sankyo, GlaxoSmithKline, Pfizer,

/Daiichi-Sankyo, Merck, and Sanofi. Dr. Dalby has

and has received personal fees from Aspen, Bayer,

rvier outside of the submitted work. Dr. Aylward has

on the advisory board of AstraZeneca. Dr. Abola has

al advisory board for AstraZeneca during the conduct

personal fees as a member of the advisory board of

r Bayer, Boehringer Ingelheim, and Daiichi-Sankyo;

m, Daiichi-Sankyo, and Population Health Research

and stockholder of AstraZeneca. Dr. Braunwald has

received research grant support through Brigham

Critical Diagnostics, Daiichi-Sankyo, Eisai, Gilead,

fi, and Takeda; and has served as a consultant for

all #$10,000/year). All other authors have reported

close.

23, 2016, accepted March 25, 2016.

TABLE 1 Baseline Characteristics

No PAD(n ¼ 20,017)

PAD(n ¼ 1,143) p Value

Age, yrs 65 (59–71) 66 (60–73) 0.0003

Age $75 yrs 2,892 (14.45) 191 (16.71) 0.035

Female 4,806 (24.01) 254 (22.22) 0.1682

Body mass index, kg/m2 27.8 (25.16–31.16) 27.8 (24.84–31.14) 0.0845

Caucasian 17,260 (86.23) 1,066 (93.26) <0.0001

History of hypertension 15,441 (77.14) 966 (84.51) <0.0001

History of hypercholesterolemia 15,317 (76.53) 924 (80.84) 0.0008

Current smoker 3,191 (15.95) 345 (30.18) <0.0001

History of COPD 444 (6.66) 68 (16.83) <0.0001

eGFR <60 ml/min/1.73 m2 (MDRD) 4,494 (22.73) 355 (31.47) <0.0001

History of diabetes 6,324 (31.59) 482 (42.17) <0.0001

History of CHF 3,888 (19.42) 338 (29.57) <0.0001

History of stroke or TIA 312 (1.56) 35 (3.06) 0.0001

History of second prior MI 3,207 (16.02) 292 (25.55) <0.0001

Multivessel coronary artery disease 11,821 (59.07) 736 (64.39) 0.0004

Prior angina 6,097 (30.46) 458 (40.07) <0.0001

Prior CABG 827 (4.13) 148 (12.95) <0.0001

History of PCI with stenting 16,028 (80.07) 863 (75.5) <0.0001

ABI #0.90 0 217 (18.99) —

Peripheral revascularization 0 389 (34.03) —

History of claudication 0 747 (65.35) —

Region <0.0001

North America 3,669 (18.33) 238 (20.82)

South America 2,342 (11.7) 116 (10.15)

Western Europe (including South Africa) 5,819 (29.07) 317 (27.73)

Eastern Europe 5,867 (29.31) 423 (37.01)

Asia/Australia/New Zealand 2,320 (11.59) 49 (4.29)

Time from qualifying MI, yrs 1.7 (1.24–2.33) 1.7 (1.23–2.35) 0.762

Qualifying MI type <0.0001

Unknown 1,142 (5.71) 81 (7.11)

STEMI 10,814 (54.09) 515 (45.18)

NSTEMI 8,037 (40.2) 544 (47.72)

Qualifying MI $2 yrs prior 7,707 (38.55) 447 (39.21) 0.6551

Time from last dose of P2Y12 inhibitor, days 152 (16–418) 86 (6.5–361.5) <0.0001

Aspirin use at baseline 19,989 (99.86) 1,142 (99.91) 0.6414

Statin use at baseline 18,537 (92.61) 1,067 (93.35) 0.3483

Beta-blocker use at baseline 16,541 (82.63) 944 (82.59) 0.9688

ACE-I or ARB use at baseline 16,093 (80.4) 936 (81.89) 0.2155

Values are median (IQR) or n (%).

ABI ¼ ankle-brachial index; ACE ¼ angiotensin-converting enzyme; ARB ¼ angiotensin receptor blocker;CABG ¼ coronary artery bypass graft; CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonarydisease; eGFR ¼ estimated glomerular filtration rate; IQR ¼ interquartile range; MDRD ¼ Modification of Diet inRenal Disease; MI ¼ myocardial infarction; NSTEMI ¼ non–ST-segment elevation myocardial infarction; PAD ¼peripheral artery disease; PCI ¼ percutaneous coronary intervention; STEMI ¼ ST-segment elevation myocardialinfarction; TIA ¼ transient ischemic attack.

J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6 Bonaca et al.J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8 Ticagrelor in Patients With PAD and Prior MI

2721

antithrombotic therapies to reduce limb morbidity inpatients with PAD.

The association of concomitant PAD with anincreased risk of ischemic events is complicated by thefact that PAD is also associated with an increased riskof bleeding (13). There are many shared comorbiditiesthat increase the risk of both PAD and bleeding,including age and renal dysfunction. In addition, it ispossible that the diseased vasculature in PAD patientsmay be more prone to injury or procedural complica-tions leading to bleeding. In trials of potent antith-rombotic strategies for prevention in PAD, there havebeen significant increases in bleeding risk, particu-larly with full-dose anticoagulants (7,8,10,14).

The PEGASUS-TIMI 54 (Prevention of Cardiovas-cular Events in Patients With Prior Heart Attack Us-ing Ticagrelor Compared to Placebo on a Backgroundof Aspirin–Thrombolysis In Myocardial Infarction 54)trial demonstrated that the addition of ticagrelor, apotent, reversible P2Y12 inhibitor, reduces long-termischemic risk in patients with a history of MI. Thisreduction in ischemic risk was accompanied by anincrease in TIMI (Thrombolysis In Myocardial Infarc-tion) major bleeding. We therefore investigated theischemic and bleeding risks associated with concomi-tant symptomatic PAD and whether the effects ofticagrelor on MACE and major adverse limb events(MALE) were modified by the presence of PAD.

METHODS

STUDY POPULATION. The PEGASUS-TIMI 54 trialrandomized patients with prior MI to ticagrelor 90 mgtwice daily, ticagrelor 60 mg twice daily, or placebo,all on a background of low-dose aspirin. The design(15) and primary results of the trial have been pub-lished (16). In brief, the trial enrolled 21,162 patientswith a spontaneous MI occurring 1 to 3 years prior toenrollment and at least 1 of the following additionalhigh-risk features: age $65 years; diabetes mellitusrequiring medication; a second prior spontaneous MI;multivessel coronary artery disease; or chronic renaldysfunction, defined as a creatinine clearance <60ml/min, as estimated by the Cockcroft-Gault equa-tion. Patients were ineligible if there was planned useof a P2Y12 receptor antagonist or anticoagulant ther-apy during the study period; if they had a bleedingdisorder, or a history of an ischemic stroke or intra-cranial bleeding, a central nervous system tumor, oran intracranial vascular abnormality; or if they hadhad gastrointestinal bleeding within the previous 6months or major surgery within the previous month.The presence of PAD was determined at baseline in21,160 patients. In addition, among patients with

PAD sites were to report whether there was anankle-brachial index (ABI) #0.90, a history of periph-eral revascularization, or a history of claudication.ENDPOINTS. The primary efficacy endpoint was thecomposite of CV death, MI, or stroke (MACE). ALIwas prospectively collected and adjudicated usinga previously published definition that requiredboth a clinical presentation consistent with acuteischemia and findings either on physical examinationor by imaging (12). Peripheral revascularization

Bonaca et al. J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6

Ticagrelor in Patients With PAD and Prior MI J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8

2722

procedures were site reported and were collectedprospectively on a dedicated electronic case reportform. A trained reviewer (M.P.B.), blinded to treat-ment assignment, identified peripheral revasculari-zation for ischemia (e.g., chronic critical limb

FIGURE 1 Outcomes in Placebo Patients by PAD at Baseline

Risk of MACE with P25%

20%

15%

10%

5%

0%0 180 360 54

Unadjusted HR 2.46 (95% CAdjusted HR 1.60 (95% CI 1

Days from Ra

CV D

eath

, MI,

or S

trok

e (%

)

Number at RiskPAD 404 384 367 34

No PAD 6663 6508 6394 615

no PAD

HR(95%CI)p–value

3.32(2.30–4.79)

P<0.001

1.93(1.35-2.75)P<0.001

16%

3.0%

5.0%

9.6% 9.5%

1.8%

4.0%

0.1%

1.0%

7.0%

12.

14%

12%

10%

8%

6%

4%

2%

0%CVD Myocardial

InfarctionStroke Acute limb

ischemiaCoronaRevasc

2.80(1.68–4.67)

P<0.001

13.70(3.67–50.90)

P<0.001

1.7(1.31–2P<0.0

N=6663

A

B

KM R

ate

at 3

Yea

rs (%

)

(A) CV death, MI, or stroke at 3 years in the placebo arm by known PAD

endpoints at 3 years in the placebo arm by known PAD at baseline. Unad

patients with PAD versus no PAD. ARR ¼ absolute risk reduction; CI ¼ c

intracranial hemorrhage; KM ¼ Kaplan-Meier; MACE ¼ major adverse ca

artery disease; TIMI ¼ Thrombosis In Myocardial Infarction.

ischemia, ALI, or claudication). The endpoint ofMALE was defined as the composite of ALI and pe-ripheral revascularization for ischemia. Amputationwas collected as a safety event and included ampu-tations for any cause (e.g., trauma, infection, or

AD vs no PAD

0 720 900 1080

I 1.92 – 3.15), p<0.001.20 – 2.13), p=0.0013

ndomization

PAD PatientsN=40419.3%

Patients without PADN=6663

8.4%

4 309 223 1049 5567 4120 1924

5%

9.2%

4.6%

1.0%1.6% 1.4%

2.2%

0.6%1.3%

0.5%

14.0%

ry.

PeripheralRevasc.

All CauseMortality

TIMI MajorBleeding

TIMIMajor/Minor

Bleeding

ICH or FatalBleeding

8.43)01

19.04(11.47–31.59)

P<0.001

3.16(2.35–4.27)

P<0.001

1.47(0.53–4.07)

P=0.46

1.67(0.72–3.85)

P=0.23

2.05(0.62–6.76)

P=0.24

PADN=404

at baseline with unadjusted and adjusted HRs. (B) Efficacy and safety

justed HRs, 95% CIs, and p values are shown for outcomes in placebo

onfidence interval; CV ¼ cardiovascular; HR ¼ hazard ratio; ICH ¼rdiovascular events; MI ¼ myocardial infarction; PAD ¼ peripheral

FIGURE 2 Risk of MACE by Type of PAD

0.1 0.75 2.5 5 101.0

2.8720.9

18.6

17.7 8.4

8.4

8.4 (1.94 – 4.26) <0.001

No PAD Group = 6,663

(1.69 – 3.27) <0.001

(0.91 – 4.06) 0.085Ankle Brachial Index ≤ 0.90N=48

ClaudicationN=222

Hazard Ratio forCVD/MI/Stroke

3 Year KMRate (%)

PAD No PAD (95% CI) P–Value

Prior PeripheralRevascularizationN=133

2.35

1.93

Risk of major adverse cardiovascular events by type of PAD versus no PAD (n ¼ 6,663) at

baseline in placebo arm. Categories were hierarchical including any history of peripheral

revascularization, claudication with no history of peripheral revascularization, and

abnormal ankle brachial index (#0.90) only with no claudication or history of peripheral

revascularization. CVD ¼ cardiovascular death; other abbreviations as in Figure 1.

FIGURE 3 Adjusted Risk of Efficacy and Safety Events by PAD Status

Adjusted* HazardRatio

3 Year KMRate (%)

PAD No PAD (95% CI) P-Value

CVD /MI / Stroke

CV Death

Myocardial Infarction

Stroke

Mortality

TIMI Major Bleeding

1.84

1.60

1.33

2.31

2.05

1.57

19.3

9.6

9.5

4.0

14.0

1.6

4.6

1.0

3.0

8.4

5.0

1.8

(1.20 – 2.13)

(1.16 – 2.94)

(0.91 – 1.94)

(1.26 – 4.25)

(1.43 – 2.94)

(0.47 – 5.22)

0.0013

0.0102

0.14

0.46

0.0071

<0.001

1.00.1 10PAD PADBetter Worse

Adjusted risk of efficacy and safety endpoints in the placebo arm by known PAD versus no

PAD at baseline. HRs, 95% CIs, and p values shown are all from the adjusted model.

*Adjusted for age, race, MI type, second prior MI, diabetes, multivessel disease, hyper-

tension, hypercholesterolemia, smoking, CHF, COPD, prior stroke/TIA, angina, CABG, PCI

with stenting, time from P2Y12 withdrawal, eGFR, and region. Abbreviations as in Figures 1

and 2.

J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6 Bonaca et al.J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8 Ticagrelor in Patients With PAD and Prior MI

2723

ischemia). Due to the lack of specificity, this outcomewas not included in the MALE endpoint for thisanalysis, but was included in a sensitivity analysis ofMALE by treatment arm. The primary safety endpointwas TIMI major bleeding. Additional safety endpointswere TIMI minor bleeding, intracranial hemorrhage(ICH), and fatal bleeding. A clinical events committeeblinded to treatment allocation adjudicated all pri-mary efficacy outcomes, ALI, and bleeding events.STATISTICAL ANALYSIS. Baseline characteristicswere summarized using medians and quartiles forcontinuous variables and frequencies and percent-ages for categorical variables. Differences were testedwith the Wilcoxon rank sum test for continuous var-iables and with the Pearson chi-square test for cate-gorical data. Cox proportional hazard models wereused to assess the risk of MACE. Associations betweenPAD and the risk of MACE were adjusted for baselineclinical characteristics that differed significantly be-tween patients with and without PAD (age, race,MI type, second prior MI, diabetes, multivessel dis-ease, hypertension, hypercholesterolemia, smoking,congestive heart failure, chronic obstructive pulmo-nary disease, prior stroke/transient ischemic attack,angina, coronary artery bypass graft [CABG], timefrom P2Y12 withdrawal, history of percutaneous cor-onary intervention with stenting, estimated glomer-ular filtration rate, and region). Clinical predictors ofMALE were evaluated in patients overall and then inthose with known PAD at baseline. Clinical baselinecharacteristics that differed (p < 0.1) were enteredinto a backward model selection to identify inde-pendent predictors with variables with p values>0.05, which were removed 1 at a time until all var-iables remaining had a p value <0.05. Analyses of theefficacy and safety of ticagrelor were not adjustedbecause treatment was randomized; therefore, base-line characteristics and potential confounders wereapproximately balanced (Online Table 1). The pro-portional hazards assumption was visually examinedby scaled Schoenfeld residual plot, and the validity ofthis assumption was not violated. Efficacy analyseswere performed on an intention-to-treat basis. Safetyanalyses included all patients who received at least 1dose of study drug and included all events occurringafter receipt of the first dose and within 7 days of thelast dose of study drug. The prognostic accuracy ofthe models was compared using the area under thecurve derived from receiver-operating characteristiccurves (C-statistic) (17). The ability of the model toenhance discrimination and correctly reclassify sub-jects was additionally tested with the integrateddiscrimination improvement and the category-freenet reclassification improvement (18,19).

RESULTS

A total of 21,162 patients were enrolled into thePEGASUS-TIMI 54 trial with prior MI and 1,143 (5%)had known PAD at randomization. Patients with

CENTRAL IL

Bonaca, M.P. et al

Efficacy of ticagrelo

reduction; CI ¼ confi

Bonaca et al. J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6

Ticagrelor in Patients With PAD and Prior MI J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8

2724

known PAD primarily had a history of claudication(65%), with approximately one-third (34%) having ahistory of a peripheral revascularization procedureand only 48 (w4%) with an abnormal ABI onlywithout prior revascularization or symptoms(Table 1). Baseline characteristics in those with andwithout PAD at baseline are shown in Table 1, and bytreatment arm within the PAD group in OnlineTable 1. The presence of PAD at baseline was associ-ated with older age, qualifying non–ST-segmentelevation MI relative to ST-segment elevation MI,diabetes, renal dysfunction, smoking status, heartfailure, history of stroke or transient ischemic attack,and time from last dose of P2Y12 inhibitor prior torandomization (Table 1). Patients with PAD at baseline

LUSTRATION Ticagrelor in Patients With PAD and Pr

. J Am Coll Cardiol. 2016;67(23):2719–28.

r in patients with PAD: CV death, MI, or stroke at 3 years with ticagrelor (po

dence interval; CV ¼ cardiovascular; HR ¼ hazard ratio; MI ¼myocardial infarc

were more likely to have multivessel coronary diseaseand a history of CABG, and less likely to have a historyof percutaneous coronary intervention (Table 1). Useof background medical therapy, including aspirin(99.9%), statins (93%), beta-blockers (83%), andangiotensin-converting enzyme inhibitors or angio-tensin receptor blockers (80.5%) was high andsimilar, regardless of known PAD.

PAD AND RISK IN PATIENTS RANDOMIZED TO

PLACEBO. At 3 years, patients with concomitant PADat baseline randomized to placebo had more than a2-fold increase in the rate of MACE events relative topatients without concomitant PAD (19.3% vs. 8.4%;unadjusted hazard ratio [HR]: 2.46; 95% confidence

ior MI: 3-Year Results

oled doses, purple) versus placebo (green). ARR ¼ absolute risk

tion; NNT ¼ number needed to treat; PAD ¼ peripheral artery disease.

J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6 Bonaca et al.J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8 Ticagrelor in Patients With PAD and Prior MI

2725

interval [CI]: 1.92 to 3.15) (Figure 1A). Rates of theindividual components of CV death, MI, or stroke, aswell as all-cause mortality were also higher in pa-tients with PAD (Figure 1B). The risk of MACE associ-ated with PAD at baseline was highest in patients witha history of prior peripheral revascularization, fol-lowed by those with claudication but no history ofrevascularization, and patients with an abnormal ABI(#0.90) alone (Figure 2). Among patients with PAD,components of MALE, including peripheral revascu-larization, occurred in 9.2% of patients and ALI in1.0%, whereas (as expected) these events occurredin <1% of patients without known PAD at randomi-zation (Figure 1B). Overall, beyond PAD history, in-dependent predictors of MALE at baseline were priorCABG, current smoking, hypertension, chronicobstructive pulmonary disease, and diabetes (OnlineTable 2). For the patients with known PAD, a priorhistory of peripheral revascularization (HR: 3.76; 95%CI: 2.26 to 6.25; p < 0.001) was the only independentpredictor of future MALE.

Patients in the placebo arm with PAD tended tohave higher rates of bleeding at 3 years comparedwith patients without known PAD, including TIMImajor bleeding (1.56% vs. 1.03%; HR: 1.47; 95% CI:0.53 to 4.07; p ¼ 0.46) (Figure 1B), TIMI major or mi-nor bleeding (2.21% vs. 1.37%; HR: 1.67; 95% CI: 0.72to 3.85; p ¼ 0.23), and the composite of ICH or fatalbleeding (1.25% vs. 0.57%; HR: 2.05; 95% CI: 0.62 to6.76, p ¼ 0.24) (Figure 1B).

After adjusting for differences in baseline charac-teristics, patients with concomitant PAD at baseline

TABLE 2 Efficacy and Safety of Ticagrelor in Patients With PAD

Placebon, %

Ticagrelor 60 mgn, %

Efficacy outcomes, N 404 368

CV death, MI, stroke 71, 19.3 47, 14.1

CV death 34, 9.6 15, 4.2

All-cause mortality 51, 14.0 25, 8.2

Stroke 17, 4.0 8, 2.9

Ischemic stroke 16, 3.7 8, 2.9

Safety outcomes, N 399 363

TIMI major 4, 1.6 4, 1.6

TIMI major or minor 6, 2.2 7, 3.2

ICH or fatal bleeding 3, 1.3 0, —

Values are n, 3-year Kaplan-Meier (%), unless otherwise indicated.

CI ¼ confidence interval; CV ¼ cardiovascular; HR ¼ hazard ratio; ICH ¼ intracranial h

had a significantly higher risk of MACE relative topatients without concomitant PAD (adjusted HR:1.60; 95% CI: 1.20 to 2.13; p ¼ 0.0013) (Figure 1A). El-ements of the model before and after the addition ofPAD, as well as the C-statistic change and measures ofreclassification, are shown in Online Tables 3 to 5.Directionally similar results existed for the compo-nents of MACE, particularly CV death (adjusted HR:1.84; 95% CI: 1.16 to 2.94; p ¼ 0.0102) and stroke(adjusted HR: 2.31; 95% CI: 1.26 to 4.25; p ¼ 0.0071).All-cause mortality was also significantly higher(adjusted HR: 2.05; 95% CI: 1.43 to 2.94; p < 0.001)(Figure 3). The adjusted risk of TIMI major bleeding(adjusted HR: 1.57; 95% CI: 0.47 to 5.22; p ¼ 0.46)(Figure 3), TIMI major or minor bleeding (adjustedHR: 1.51; 95% CI: 0.53 to 4.30; p ¼ 0.44), and thecomposite of ICH or fatal bleeding (adjusted HR: 1.97;95% CI: 0.44 to 8.80; p ¼ 0.37) tended to remainhigher in patients with PAD.

EFFICACY AND SAFETY OF TICAGRELOR FOR MACE

IN PATIENTS WITH PRIOR MI AND PAD. There was noheterogeneity in the relative risk reduction withticagrelor for MACE in patients on the basis of thepresence of PAD at baseline (p interaction ¼ 0.41)(Central Illustration). However, by nature of theirgreater absolute risk, patients with PAD had anumerically greater absolute risk reduction at 3 years(4.1%; 95% CI: �1.07% to 9.29%) relative to thosewithout PAD (1.0%; 95% CI: 0.14% to 1.9%) (CentralIllustration). The efficacy versus placebo was gener-ally similar for the 2 doses of ticagrelor, but, of note,

Ticagrelor 90 mgn, %

Ticagrelor 60 mgHR (95% CI)

p Value

Ticagrelor 90 mgHR (95% CI)

p Value

371

54, 16.3 0.69 (0.47–0.99)p ¼ 0.045

0.81 (0.57–1.15)p ¼ 0.24

26, 7.9 0.47 (0.25–0.86)p ¼ 0.014

0.83 (0.50–1.38)p ¼ 0.46

41, 11.7 0.52 (0.32–0.84)p ¼ 0.0074

0.88 (0.58–1.32)p ¼ 0.53

10, 3.1 0.49 (0.21–1.14)p ¼ 0.097

0.63 (0.29–1.38)p ¼ 0.25

7, 2.0 0.52 (0.22–1.22)p ¼ 0.13

0.47 (0.19–1.14)p ¼ 0.095

368

5, 1.8 1.18 (0.29–4.70)p ¼ 0.82

1.46 (0.39–5.43)p ¼ 0.57

8, 3.1 1.36 (0.46–4.05)p ¼ 0.58

1.57 (0.54–4.53)p ¼ 0.40

1, 0.4 — 0.39 (0.04–3.75)p ¼ 0.42

emorrhage; MI ¼ myocardial infarction; TIMI ¼ Thrombosis In Myocardial infarction.

FIGURE 4 Efficacy and Safety of Ticagrelor in Patients With PAD

CVD / MI / Stroke

CV Death

Stroke

Mortality

TIMI Major Bleeding

Myocardial Infarction

0.11.0

10Ticagrelor Better Placebo Better

Absolute RiskDifference at

3 Years HR (95% CI) P-Value

Ticagrelor 90 mgTicagrelor 60 mg

– 3.0– 5.2

– 1.7– 5.4

– 2.0– 1.0

– 0.9– 1.1

– 2.3– 5.7

0.220.02

0.81 (0.57 – 1.15)0.69 (0.47 – 0.99)

0.83 (0.50 – 1.38)0.47 (0.25 – 0.86)

0.76 (0.45 – 1.28)0.87 (0.53 – 1.44)

0.63 (0.29 – 1.38)0.49 (0.21 – 1.14)

0.88 (0.58 – 1.32)0.52 (0.32 – 0.84)

1.46 (0.39 – 5.43)1.18 (0.29 – 4.70)

0.0450.24

0.0140.46

0.590.30

0.0970.25

0.820.57

0.00740.53

Efficacy and safety of ticagrelor 60 mg twice daily (purple) and ticagrelor 90 mg twice

daily (green) versus placebo in patients with PAD. Abbreviations as in Figures 1 and 2.

FIGURE 5 Ticagrel

Acut

e Li

mb

Isch

emia

or

Perip

hera

l Rev

ascu

lariz

atio

n fo

r Isc

hem

ia (%

)

1.0%

0.8%

0.6%

0.4%

0.2%

0.0%

Number at RiskPlacebo

Ticagrelor

0

706714095

MALE at 3 years wit

major adverse limb e

Bonaca et al. J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6

Ticagrelor in Patients With PAD and Prior MI J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8

2726

the 60-mg dose resulted in a significant reduction inCV mortality (HR: 0.47; 95% CI: 0.25 to 0.86;p ¼ 0.014) (Table 2, Figure 4), which drove a reductionin all-cause mortality (HR: 0.52; 95% CI: 0.32 to 0.84;p ¼ 0.0074).

Ticagrelor increased TIMI major bleeding consis-tently for patients with PAD (pooled doses HR: 1.32;95% CI: 0.41 to 4.29) and without PAD (HR: 2.59; 95%CI: 1.91 to 3.52; p interaction ¼ 0.28). Although pa-tients with PAD randomized to placebo were atgreater absolute risk of bleeding, the absolute risk

or and MALE in Patients With PAD

Placebo Ticagrelor Pooled

0.71%

0.46%

HR 0.6595% CI (0.44 – 0.95)

P=0.026

180 360 540 720 900 1080

698813929

691213789

670113425

607712186

45189154

21234296

Days from Randomization

h ticagrelor (pooled doses, purple) versus placebo (green). MALE ¼vents; other abbreviations as in Figure 1.

increase with ticagrelor at 3 years was numericallylower for those with PAD (0.12%; 95% CI: �1.79% to2.04%) relative to those without PAD (1.46%; 95%CI: 1.01% to 1.91%) at baseline, although events wereinfrequent. Secondary bleeding outcomes by baselinePAD for both ticagrelor doses are shown in Table 2.Rates of intracranial hemorrhage and fatal bleeding inpatients with PAD were low.

MALE WITH TICAGRELOR. Overall, ticagrelorreduced MALE (HR: 0.65; 95% CI: 0.44 to 0.95;p ¼ 0.026) (Figure 5), with a numerically greaterrelative risk reduction with the 90-mg dose (HR: 0.49;95% CI: 0.30 to 0.81; p ¼ 0.005) relative to the 60-mgdose (HR: 0.81; 95% CI: 0.53 to 1.24; p ¼ 0.33)(Table 3). The effects were consistent for both areduction in ALI (pooled doses HR: 0.56; 95% CI: 0.23to 1.37) and peripheral revascularization for limbischemia (pooled doses HR: 0.63; 95% CI: 0.43 to 0.93)(Online Figure 1). Results were similar in thesubgroup of patients with a known history of PAD(Online Table 6) as well as when including amputa-tion for any reason (Online Table 7).

DISCUSSION

This study demonstrates that in outpatients in stablecondition with a history of MI, concomitant PADis associated with significantly heightened risk ofsystemic ischemic events, limb ischemic events,bleeding, and all-cause mortality. In addition, thebenefit of ticagrelor for the relative risk reduction ofMACE was consistent, regardless of the presence orabsence of known PAD; however, by nature of theirheightened risk, patients with PAD had a particularlyrobust absolute risk reduction (Central Illustration).Finally, the combination of aspirin and ticagrelorreduced MALE relative to aspirin monotherapy.

The association of PAD with increased MACE risk inpatients who are already at heightened risk by natureof a prior MI is complex. Possible explanationsinclude that the presence of disease in more than 1bed indicates a more severe form of atheroscleroticvascular disease. In addition, PAD is associated withother markers of atherothrombotic risk, includingage, renal dysfunction, diabetes, and smoking. Afteradjustment for these baseline differences, however,PAD remained a potent marker of systemic ischemicrisk. Regardless of whether PAD is a marker of a moresevere vascular phenotype, an integrator of multipleadverse risk factors, or both, in practice it is an easilyidentifiable and potent indicator of future risk. In thecurrent study, the magnitude of risk for MACE asso-ciated with claudication was similar to that of ahistory of peripheral revascularization, providing

TABLE 3 Limb Vascular Efficacy With Ticagrelor

Placebo(n ¼ 7,067)

n, %

Ticagrelor 60 mg(n ¼ 7,045)

n, %

Ticagrelor 90 mg(n ¼ 7,050)

n, %

Ticagrelor 60 mgHR (95% CI)

p Value

Ticagrelor 90 mgHR (95% CI)

p Value

Acute limb ischemia or peripheral revascularizationfor ischemia

47, 0.71 38, 0.60 23, 0.32 0.81 (0.53–1.24)p ¼ 0.33

0.49 (0.30–0.81)p ¼ 0.005

Acute limb ischemia 9, 0.13 6, 0.11 4, 0.06 0.67 (0.24–1.87) 0.45 (0.14–1.45)

Peripheral revascularization for ischemia 46, 0.70 37, 0.59 21, 0.29 0.80 (0.52–1.24) 0.46 (0.27–0.76)

Peripheral revascularization 60, 0.94 46, 0.74 38, 0.55 0.77 (0.52–1.13) 0.63 (0.42–0.95)

Values are n, 3-year Kaplan-Meier (%), unless otherwise indicated.

Abbreviations as in Table 2.

J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6 Bonaca et al.J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8 Ticagrelor in Patients With PAD and Prior MI

2727

validity to this clinical indicator of PAD. These find-ings underscore the importance of identifyingcomorbid PAD in the clinic to inform future risks ofboth MACE and MALE.

Identification of heightened MACE risk is particu-larly useful if it informs clinical decision-making. ThePEGASUS-TIMI 54 trial demonstrated that ticagreloris efficacious for the reduction of MACE in patientswith prior MI. Still, the need to balance ischemic ef-ficacy and bleeding risk has created a desire toexplore subgroups where ticagrelor has particularlyrobust efficacy. In the current analysis, we demon-strate that the presence of PAD does not modifythe relative efficacy of ticagrelor. However, due totheir more than 2-fold higher MACE risk, patientswith PAD enjoyed a robust absolute risk reduction of4.1% at 3 years, translating into a number needed totreat (NNT) of 25 (annualized NNT ¼ 74). Althougheach dose was tested independently against placebo,and there was no formal interaction between the ef-ficacy of either dose and the baseline presence ofknown PAD, the clearest benefit occurred with the60-mg dose, with a 5.2% absolute risk reduction inMACE and NNT of 20 (annualized NNT ¼ 58), as wellas significant reductions in both CV death and all-cause mortality. At the same time, bleeding withticagrelor appeared to be similar in patients with andwithout PAD. Therefore, the presence of PAD may behelpful to clinicians in identifying patients with priorMI who are likely to have a robust benefit withticagrelor.

In addition to MACE risk, patients with PAD in thePEGASUS-TIMI 54 trial were at heightened risk ofMALE. Limb revascularization for ischemia was asfrequent as MI and occurred at more than twice therate of stroke in patients with PAD. A history ofperipheral revascularization was a predictor of MALEin patients with PAD in this analysis, corroboratingobservations from other studies (11).

To date, dual antiplatelet therapy with aspirin andclopidogrel has not been prospectively shown to be

efficacious for reducing limb vascular events in pa-tients with PAD (7). Although not powered to addressthis question, the current study suggests that dualantiplatelet therapy with aspirin and ticagrelor maybe efficacious for reducing ALI if studied in larger,higher-risk cohorts. In addition, the reduction in theexploratory endpoint of peripheral revascularizationfor ischemia also supports potential antithromboticbenefits of ticagrelor on the limb circulation. Whetherpotent antiplatelet therapy can reduce elective revas-cularization for claudication, as has been observedwith other mechanisms, remains unknown (11). Thesequestions will be evaluated as secondary outcomes inthe ongoing EUCLID (A Study Comparing Cardiovas-cular Effects of Ticagrelor and Clopidogrel in PatientsWith Peripheral Artery Disease [NCT01732822]) trial,which is comparing ticagrelor and clopidogrel asmonotherapy for a primary endpoint of MACE.STUDY LIMITATIONS. The PAD cohort, althoughcomprising more than 1,000 patients, was a subgroupof the PEGASUS-TIMI 54 trial and thus had morelimited power for exploration of the more rareendpoints, such as ALI. In addition, patients wereidentified as having PAD at baseline by study in-vestigators, but no formal screening was required,and some patients may therefore have entered thetrial with undiagnosed PAD. Questionnaires toimprove the ascertainment and enable adjudicationof claudication as a marker of PAD were not used,thereby potentially decreasing the specificity of thisbaseline characteristic. Amputations were collectedas safety data and were not adjudicated, therebylimiting ascertainment and specificity of thisoutcome. In addition, details regarding procedurestypically considered in the endpoint of MALE (e.g.,thrombolysis/thrombectomy) were not available,underscoring the importance of examining theseoutcomes in dedicated studies in patients withPAD. Finally, peripheral revascularizations weresite-reported; however, they were collected on aspecific case report form page with instructions.

PERSPECTIVES

COMPETENCY IN PATIENT CARE: Prolonged

therapy with ticagrelor reduced ischemic risk and

increased bleeding in patients with prior MI, and in a

subgroup analysis, those with concomitant PAD

appeared to derive greater absolute benefit than

those without PAD.

TRANSLATIONAL OUTLOOK: An ongoing pro-

spective trial of ticagrelor as antiplatelet monotherapy

should providemore insight into the efficacy and safety

of ticagrelor monotherapy versus clopidogrel mono-

therapy in a broad population of patients with PAD,

including those without concomitant coronary disease.

Bonaca et al. J A C C V O L . 6 7 , N O . 2 3 , 2 0 1 6

Ticagrelor in Patients With PAD and Prior MI J U N E 1 4 , 2 0 1 6 : 2 7 1 9 – 2 8

2728

CONCLUSIONS

In patients with prior MI, concomitant PAD is asso-ciated with heightened risk of ischemic vascularcomplications and mortality, even after adjusting fortheir extensive comorbidities. In these patients, thecombination of ticagrelor and aspirin for long-termsecondary prevention appeared to reduce ischemicrisk with a particularly robust absolute risk reduction,including reductions in MACE and limb vascularevents.

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Marc P. Bonaca, TIMI Study Group, CardiovascularDivision, Brigham and Women’s Hospital, 75 FrancisStreet, Boston, Massachusetts 02115. E-mail:mbonaca@partners.org.

RE F E RENCE S

1. Bhatt DL, Steg PG, Ohman EM, et al., forthe REACH Registry Investigators. Internationalprevalence, recognition, and treatment of cardio-vascular risk factors in outpatients with athero-thrombosis. JAMA 2006;295:180–9.

2. Bhatt DL, Eagle KA, Ohman EM, et al., for theREACH Registry Investigators. Comparative de-terminants of 4-year cardiovascular event rates instable outpatients at risk of or with athero-thrombosis. JAMA 2010;304:1350–7.

3. Steg PG, Bhatt DL, Wilson PW, et al., for theREACH Registry Investigators. One-year cardio-vascular event rates in outpatients with athero-thrombosis. JAMA 2007;297:1197–206.

4. Antithrombotic Trialists’ Collaboration. Collab-orative meta-analysis of randomised trials ofantiplatelet therapy for prevention of death,myocardial infarction, and stroke in high riskpatients. BMJ 2002;324:71–86.

5. CAPRIE Steering Committee. A randomised,blinded, trial of clopidogrel versus aspirin in pa-tients at risk of ischaemic events (CAPRIE). Lancet1996;348:1329–39.

6. Janzon L, Bergqvist D, Boberg J, et al. Pre-vention of myocardial infarction and stroke inpatients with intermittent claudication; effects ofticlopidine. Results from STIMS, the SwedishTiclopidine Multicentre Study. J Intern Med 1990;227:301–8.

7. Belch JJF, Dormandy J, for the CASPAR WritingCommittee. Results of the randomized, placebo-controlled clopidogrel and acetylsalicylic acid inbypass surgery for peripheral arterial disease(CASPAR) trial. J Vasc Surg 2010;52:825–33, 833.e1–2.

8. Bhatt DL, Flather MD, Hacke W, et al., for theCHARISMA Investigators. Patients with priormyocardial infarction, stroke, or symptomatic pe-ripheral arterial disease in the CHARISMA trial.J Am Coll Cardiol 2007;49:1982–8.

9. Cacoub PP, Bhatt DL, Steg PG, et al., for theCHARISMA Investigators. Patients with peripheralarterial disease in the CHARISMA trial. Eur Heart J2009;30:192–201.

10. Magnani G, Bonaca MP, Braunwald E, et al.Efficacy and safety of vorapaxar as approved forclinical use in the United States. J Am Heart Assoc2015;4:e001505.

11. Bonaca MP, Scirica BM, Creager MA, et al.Vorapaxar in patients with peripheral artery dis-ease: results from TRA2�P-TIMI 50. Circulation2013;127:1522–9, 1529.e1–6.

12. BonacaMP,Gutierrez JA,CreagerMA,etal. Acutelimb ischemia and outcomes with vorapaxar in pa-tientswith peripheral artery disease: results from theTrial toAssess the Effects of Vorapaxar in PreventingHeart Attack and Stroke in Patients With Athero-sclerosis–Thrombolysis In Myocardial Infarction 50(TRA2�P-TIMI 50). Circulation 2016;133:997–1005.

13. Bhatt DL, Peterson ED, Harrington RA, et al.,CRUSADE Investigators. Prior polyvascular dis-ease: risk factor for adverse ischaemic outcomes inacute coronary syndromes. Eur Heart J 2009;30:1195–202.

14. Warfarin Antiplatelet Vascular Evaluation TrialInvestigators. Oral anticoagulant and antiplatelettherapy and peripheral arterial disease. N Engl JMed 2007;357:217–27.

15. Bonaca MP, Bhatt DL, Braunwald E, et al.Design and rationale for the Prevention of

Cardiovascular Events in Patients With Prior HeartAttack Using Ticagrelor Compared to Placebo on aBackground of Aspirin–Thrombolysis In MyocardialInfarction 54 (PEGASUS-TIMI 54) trial. Am Heart J2014;167:437–44.e5.

16. Bonaca MP, Bhatt DL, Cohen M, et al., for thePEGASUS-TIMI 54 Steering Committee and In-vestigators. Long-term use of ticagrelor in pa-tients with prior myocardial infarction. N Engl JMed 2015;372:1791–800.

17. DeLong ER, DeLong DM, Clarke-Pearson DL.Comparing the areas under two or more correlatedreceiver operating characteristic curves: anonparametric approach. Biometrics 1988;44:837–45.

18. PencinaMJ, D’Agostino RB Sr., D’Agostino RB Jr.,et al. Evaluating the added predictive ability of a newmarker: from area under the ROC curve to reclassifi-cation and beyond. Stat Med 2008;27:157–72, dis-cussion 207–12.

19. Pencina MJ, D’Agostino RB Sr., Demler OV.Novel metrics for evaluating improvement indiscrimination: net reclassification and integrateddiscrimination improvement for normal variablesand nested models. Stat Med 2012;31:101–13.

KEY WORDS acute limb ischemia, majoradverse cardiovascular events, major adverselimb events, peripheral arterial disease,ticagrelor

APPENDIX For supplemental tables and afigure, please see the online version of thisarticle.